Gibberellin structure and function: Biological activity and competitive inhibition of gibberellin 2- and 3-oxidases

Some gibberellin (GA) analogues, especially with C-16,17 modifications of GA5, can inhibit growth of plants apparently by acting as competitors with the endogenous substrate of GA biosynthetic enzymes. Here, we directly confirm the competitive action of GA derivatives but also show that some analogues may retain significant bioactivity. A recombinant 3-oxidase from pea, which converts GA20 to bioactive GA1, was inhibited by GA5, and 16,17-dihydro-GA5 derivatives, especially if the C-17 alkyl chain length was increased by up to three carbons or if the C-13 hydroxyl was acetylated. Genetic confirmation that GA5 analogues target 3-oxidases in vivo was provided by comparing the growth response of a WT (LE) pea with a 3-oxidase mutant (Ie-1). Two pea 2-oxidases that inactivate bioactive GAs, were inhibited by GA1 and GA3 but were generally insensitive to GA5 analogues. α-Amylase production by barley half-seeds in response to GA analogues provided a method to study their action when effects on GA biosynthesis were excluded. This bioactivity assay showed that 16,17-dihydro GA5 analogues have some inherent activity but mostly less than for GA5 (5-50-fold), which in turn was 100-fold less active than GA1 and GA3. However, although C-17 alkyl derivatives with one or two added carbons showed little bioactivity and were purely 3-oxidase inhibitors, adding a third carbon (the 17-n-propyl-16,17-dihydro GA5 analogue) restored bioactivity to that of GA5. Furthermore, this analogue has lost its capacity to inhibit stem elongation of Lolium temulentum (Mander et al., Phytochemistry 49:1509-1515, 1998a), although it strongly inhibits the 3-oxidase. Thus, the effectiveness of a GA derivative as a growth retardant will reflect the balance between its bioactivity and its capacity to inhibit the terminal enzyme of GA biosynthesis. The weaker growth inhibition in dicots including pea (approximately 10%) than in monocots such as L. temulentum (>35%) is suggestive of taxonomic differences in the bioactivity of GAs and/or their effects on GA biosynthesis